Now you can take a closer look at an individual state, and what comprises its score. Let's look at Ohio, the fourth-highest ranking state in risk-related impacts. We know from looking at the first sorted table in Tutorial 1 that fugitive air releases are responsible for the highest risk-related impacts in Ohio. But what chemicals are mostly responsible for these impacts?
Step 2.1 Perform a State-Level Selection for Ohio
First, limit your selected set to releases from facilities located in Ohio by performing a new selection. Click on the Select button. The screen should display your last selection, which was 'Submission.Year is equal to 2002'. To add another condition statement, click on the '1' to the left of 'Submission', then click 'Add Condition'. A new text line will appear. Click on 'Chemical Flags.Year Chemical Added', then click on 'Facility Location', then 'State'. In the blank at the end of the line, type in the abbreviation for Ohio, 'OH'. Remember that fields for which you must type the selection criteria are case sensitive. Click somewhere else in the window aside from the box you just typed in, to enter your change (all of the text in the line will then turn blue). Your selection statement should look like the one below. Click Submit. The model may take a few minutes to complete the task.
When the model is done with the selection, the Select elements... screen will disappear. The number of facilities, chemical releases, and elements that are selected in your set will be displayed in the top right corner of the screen.
Step 2.2 Results for Ohio, Chemical by Media
Click on the Custom Tables button at the end of the second row of menu buttons at the top of the screen. Now we will look at what chemicals are being released by facilities in Ohio, and the media into which they are released. Click New Table, and select 'Chemical.Chemical', and 'Release.Media Text'. The selections for your previous table will still be displayed, so make sure to deselect them. Similarly, the name of the previous table will be displayed in the 'Name' box at the bottom of the screen. Delete that name, and type in a new name for this table, for instance, 'Ohio 2002 Chemical by Media'. Hit Run! and the model will create your new table. The table may take a few minutes to generate; when it does you will see the following screen (if the graph is still displaying, click on the Table button):
Crosstab Table, Chemical by Media for Ohio, 2002
If you do not see the rows, click on the plus sign in the leftmost cell to expand the row dimension. Each cell of this table shows the risk-related score for each chemical-medium combination in Ohio.
To get an idea of the relative contribution of each combination, click on Sorted Table. This table shows the individual and cumulative contribution of each chemical-medium combination to the total impact in Ohio for 2002, in descending order.
Sorted Table, Chemical by Media for Ohio, 2002
You can see from this table that over 35 percent of the calculated score in Ohio for 2002 is due to fugitive air releases of Manganese and manganese compounds. To look at the rankings only by chemical, go back to the Table screen and collapse the rows by clicking on the minus sign to the left of 'Media Text', and then click again on the Sorted Table button, as we did in Tutorial 1. The table is shown below. You can see that, when releases to all media are considered, Manganese and manganese compounds has the highest risk-related results, and Lead and lead compounds has the second-highest ranked risk-related results.
Sorted Table, Chemical Only for Ohio, 2002
While in the Sorted Table, you can change the selection in the 'Value Selected' drop-down box, and see the rankings of chemicals by TRI Pounds, Hazard, or other model results. You will notice that the rankings can change quite dramatically. The change can be for a number of reasons. One very common reason is illustrated by the releases for Zinc and zinc compounds, which ranks first in TRI Pounds, but 37th in the Risk-related Results perspective. The reason for this is that approximately 104 million pounds (99 percent of the 105 million pounds in total) of the Zinc and zinc compounds releases are transferred to off-site recycling or disposal facilities or landfilled. The RSEI model does not consider the risk (if any) that may result from these pathways, so these releases are not reflected in the ranking by Risk-related Results.
When analyzing RSEI results, another factor to keep in mind is that while RSEI uses the best available data, inevitably some data sets will contain errors, and some assumptions are made in the absence of sufficient data. For instance, in this selection, the top chemical-medium combination by risk-related impact is fugitive air releases of Manganese and manganese compounds. This could simply reflect releases that are resulting in high risk-related impact. However, the results could also be affected by some of the data used or the assumptions that are made in modeling the water pathways, such as tthe air pathway, such as average weather patterns, population placement, etc. The third-highest combination is POTW releases of Lead and lead compounds, which may be affected by the assumptions used in modeling the water pathway, such as the locations of various points used in the exposure modeling: the facility effluent pipe, drinking water intakes, or the number of people exposed through drinking water or fishing. RSEI results are screening level only, and should be followed up with further analysis.
Step 2.3 Time Trend Analysis for High-Ranking Chemicals
You might want to see the trend in the releases of Ohio's high-ranking chemicals over several years, to see if releases and scores are increasing or decreasing. To do this, begin by modifying the last selection statement to limit the selected set to releases of these chemicals only. Click on the Select button to return to the Select elements... screen. The two statements used in the previous selection will still be showing. Click on the circle to left of the first statement, and then click on 'Add Bracket'. The text line, 'all of the following apply' will appear. This is a bracket statement telling the model how to interpret the list of conditions that will follow. We are going to add a list of four chemicals, and we want releases for any of them to be selected, so click on 'all' and then click on 'any' in the drop down list. The text line should now say 'any of the following apply'.
You will now add your list of chemicals. Click on the text, 'Chemical Flags.Year Chemical Added', and then click on 'Chemical Identifiers', then 'Chemical'. 'Chemical Identifiers.Chemical' will show in the text line. Click on the blank at the end of the text line, and a screen will pop up that lists all of the chemicals included in the RSEI model. Scroll through the list until you come to the first chemical to select, 'Manganese and manganese compounds'. Click on that chemical, and it will appear in the text line in your selection statement. Click on the '3.1' at the beginning of the line, and click 'Add Condition'. Repeat these steps to select 'Lead and lead compounds', 'Sulfuric acid', 'Chromium and chromium compounds', and 'Nickel and nickel compounds'.
In order to look at time trends, you need to make another modification to your selection statement. In the first condition statement, which says 'Submission.Year is equal to 2002', click on the 'is equal to'. Click on 'is between' in the drop down menu. The text line should now say 'Submission.Year is between 2002 and '. Click on '2002' and select '1996' from the list. Then click on the last blank and select '2006' from the list.The model will now select any releases of the four chemicals in Ohio in the years 1996 through 2006. Note that the 'is between' operator is inclusive.
Because you are doing a time trend, it is important not to inadvertently introduce other factors into your analysis. For Reporting Year 1998, TRI added a number of new industries that had not previously been required to report to TRI. If these facilities are not accounted for in the time trend analysis, the results for 1998 and after will look much higher than those for previous years, simply because more facilities are included. If you exclude the new reporters, you will be working with the same set of facilities for all years, and so will get a more accurate sense of the trend over time.
To exclude these facilities, create a new bracket statement by clicking on the first empty circle, and clicking on 'Add Bracket'. In the new line that appears, change the 'all' to 'none'. Click on the first part of the new condition statement, and select 'Facility Industry', then 'SIC Code 1'. Click on the blank space at the end of the line, and in the window that appears, select the first code, 1021 [Copper Ores]. Then add a new condition by clicking on the '4.1' at the beginning of the line, and click on 'Add Condition'. Then repeat the same steps, but instead selecting 1031 in the SIC code window. Do this for each of the following SIC codes: 1041, 1044, 1061, 1099, 1221, 1222, 1231, 4911, 4931, 4939, 4953, 5169, 5171, and 7389. Your selection statement should look like the one below.
Selection Statement for Time Trend Analysis, Ohio Releases 1996-2006
Because this is an intricate and useful query, it is a good idea to save it for later use. Type in a description, such as 'Ohio, 1996-2006 top 5 chemicals excluding new reporters', in the white box at the bottom of the window. This is a text description that will be saved with your query. Click on 'Save' at the top of the window, and enter a shorter name for your query, such as 'Ohio top 5 96-2006 exc new'. The model will automatically add a '.qry' extension to your selection name. In the future you can load this query and either resubmit it, or use it as the basis for building new selection statements.
Click Submit, and the model will perform the selection. The model may take a few minutes to complete the task. When the model is done with the selection, the Select elements... screen will disappear. The number of facilities, chemical releases, and elements that are selected in your set will be displayed in the top right corner of the screen.
One of the quickest ways to look at a trend over time is to use the preformatted graphs provided. Click on the Summary button in the second row of menu buttons, then Total by Year. This graph adds together all of the pounds and the scores of all of the chemical releases in the selected set and presents them separately for each year in the set. In this case, it would be all of the chemical releases of Manganese and manganese compounds, Lead and lead compounds, Sulfuric acid, Chromium and chromium compounds, and Nickel and nickel compounds in Ohio, excluding the new reporters. The graph is shown below. Pounds (as reported in TRI) are shown in green, and correspond to the axis on the right side of the graph; the total score is shown in red and corresponds to the axis on the left side of the graph. You can see that both pounds and score have fluctuated over time. The score was at its highest point in 2000, and has been substantially lower over the period 2001-2006 than over 1996-2000.
Total by Year Graph for Selected Chemicals in Ohio, 1996-2006
To try to identify the reason for the increased score in 2000, you can create a new custom table by chemical by media by year. Click on the Custom Tables button, and then New Table. In the list of dimensions, select 'Chemical.Chemical', 'Submission.Year', and 'Release.Media Text'. Make sure to deselect any selections from previous runs. In the 'Name' box, type 'Ohio 1996-2006 Chemical by Media by Year'. Click Run! and your table will generate (this may take a few minutes). Once the query is finished, click on Table. The table should look like the one shown below. If you do not see media listed, click on the plus sign to the right of 'Year', and the rows showing media will be expanded.
Crosstab Table, Chemical by Media by Year for
Top 5 Chemicals in Ohio, 1996-2006
Move the row headers so that 'Media Text' is to the right of 'Year', if it is not already. To move a header, click and drag the header name to the desired position, as shown in the picture below. Now collapse the media rows by clicking in the minus sign to the left of its name. You can also collapse the chemical column so that the table is shown by year only. Looking at the totals by year only, you can see that from 2000 to 2001 the risk-related score decreased by approximately 300,000. If you now expand the column showing chemical (by clicking on the plus sign), you can further identify the reason for the decrease.
Crosstab Table, Demonstrating Moving and Collapsing of Rows or Columns
To make it easier, you can use the filter to show only releases for 2000 and 2001. Click on Filter, then click on 'Year', then select 2000 and 2001. You can see that, while all five chemicals declined somewhat, Manganese and manganese compounds declined by approximately 260,000, which accounts for the great majority of the decrease overall from 2000 to 2001. Chromium and chromium compounds also decreased substantially, by approximately 30,000. If you now expand the media rows, you can see that the decrease in risk-related score is due to both stack and fugitive air releases for both chemicals.
You can also export this table, in order to do calculations in a spreadsheet, for example. While viewing your table, simply click on Export in the top row of menu buttons, and select the type of file you would like your table to be exported to. Click on the folder icon to the right of the 'Export to File' box at the bottom of the window, and select the directory in which you would like to store your exported file. The default directory for stored files is "C:\Program Files\RSEI\User," but you can export the file anywhere. Enter a name at the bottom of the screen and click Save. In the Setup of the data export screen, click OK, and the model will export your table. You can then open up the file in whatever program you selected. Your exported table will contain all of the results fields (TRI Pounds, Hazard, Risk-related Results, etc.) in one table.
The Table, Graph and Sorted Table functions will always display results based on the last crosstab table that was generated. Even if you perform a new selection using the Select button, these functions will not change until you create a new table based on your new selection.
Step 2.4 Further Analyses
If you wish to determine what facilities released Manganese and manganese compounds and Chromium and chromium compounds in 2000, you could go back to the Select elements...screen and add a condition statement by clicking on the first empty circle. Modify the condition to read, 'Release.Media Code is less than or equal to 2'. This will change your selection to only fugitive (media=1) and stack (media=2) air releases. You could also modify the first statementto be 'Submission.Year is equal to 2000'. Then you could create a new Custom Table showing 'Facility Name' by 'Year'. This would show you which facilities released Manganese and manganese compounds and Chromium and chromium compounds as stack or fugitive airreleases in 2000.
The last tutorial examines facility-level information and explains how to use some of the facility-specific features, including geographic information system (GIS) capabilities.